D1.1 Partial Pen vs. Complete Pen question

[RobsVette]

Ok Guys, I have a question and am hoping you guys can help. The question results from me doing something stupid, I want to see what you guys think. If I ultimately have to redo the weld joint I am willing to take responsibility for my mistakes and accept that, however I want to make sure I have a full understanding of what is required before I make a blanket statement that I need to redo a weld.

Here is the issue, we are welding a 12" x 20" piece of hollow stuctural steel (5/8" thick) to a 2" thick plate that is 4' x 4' to create a pipe support for a steam line. The design engineer for the structure is saying that in order to carry the loads that were determined by the stress analysis, we need to be able to have a full penetration weld. (as a disclaimer: this individual by his own admission is not a welding engineer. His position is the he wants to be absolutely certain that he can count on 100% of the cross section of the tube as load carrying. ie... there are no break on discontinuities to create stess concentrations along the weld joint)

According to AWS D1.1 this is a T or Corner joint. When reviewing the pre-qualified WPSs in D1.1 none of them allow for a simple open butt joint with a 6010 root. The only thing that I can come up with is either a TC-U4A joint (which would require backing) or a BTC-P4 joint (with a 1/8" land.) This is the rout we went. Rather than try to secure backing strips I felt it would be easier to weld with the land as a partial penetration joint.

Given the very small land used (1/8") is it still acceptable to take credit for the entire 5/8" as being welded from the tube steel to the plate. My opinion would be yes, because as you weld the first passes (7018 rod was used to weld this,) you would have a good amount of penetration down into the base metal in order to join them together. I know 7018 is considered a moderate penetration rod, but I think you would still break down the metal along the land and create a good fusion along the root of the joint. I dont think it would be a 100% "X Ray quality" weld, but I do think it would acceptable in regards to the load carrying capacity of the support.

What would you guys consider on a joint like this? Would the partial penetration be acceptable or should I go back and redo the joint to make it a complete penetration. Any refences in AWS D1.1 would be helpful as well.

Please let me know if any further information is needed.

Thanks in advance for any help.

Rob

 

[connectegr]

No a partial penetration weld is not a sufficient alternative, for the full penetration weld. As you noted the 1/8" land would is expected to provide an area of incomplete fusion. This loss of area and the reduced design strength would provide a weld insufficient to develop the full strength of the material thickness. If the condition was a plate and a TC-U4b CJP weld was used the prep would have been of the full thickness and the root pass would have been backguoged to sound metal and re-welded to provide the full thickness weld. In your case a backing bar is required, TC-U4a, because access is not available to repair the root pass. Therefore the root pass is provided on the backing base metal. This should allow for complete fusion of the material thickness. In either case of the CJP welds, UT should be performed to verify that no cracks or defects exist through the thickness of the weld.

Also note that using E7018 stick electrode the BTC-P4 weld is not "GF", and an additional 1/8" is lost to determine the effective weld strength. For a given bevel dimension, "S", the effective weld "E" = S - 1/8". Plus a minimum land of 1/8", for a minimum loss of 1/4".

If you wish to avoid full penetration welding, and the full tensile strength of the member is required... an alternative is to increase the material thickness above the design minimum required. Use of GMAW or an FCAW process can also reduce the minimum loss to 1/8", since S = E for the GF weld.

http://www.FerrellEngineering.com

 

[connectegr]

I guess another option is to UT your existing weld. Although this would not be a pre-qualified CJP weld without backing or backguoging, you might be lucky. I have seen good welders with the right settings make a clean partial penetration weld for the full thickness. However, even the best cannot do it consistently with SMAW process. If you have blow-out through the land thickness or defects, I would suggest repairing to the appropriate TC-U4a with backing.

http://www.FerrellEngineering.com

 

[RobsVette]

connectegr;

thank you for the quick response. I do appreciate it and I think you are correct. I probably need to redo this weld a make it a full penetration weld.

Could you perhaps answer a simple question;

Why are you not able to use 7018 on a open butt weld? And then further why do the prequalified AWS D1.1 procedures not involve only 7018 and not any 6010 procedures?

I assume the 7018 has something to do with the only moderate penetration characteristics of the rod. Thus if you have only moderate penetration around an open root, it will not fill in very well. Thus the requirement to use either backgouge and grind out and add a weld pass on the back of the metal or use a 6010 or equivalent rod with a much deeper penetration. After the initial root pass you then switch to the low hydrogen electrode because it produces the cleanest weld with the least number of discontinuities.

The joint we did, does have a 3/16" cap on top of it, however I dont think that adds to the effective cross section of the weld as far as joining the two metals together.

Again thank you for the quick response and advice.

 

[connectegr]

I am a little confused...

An E7018 is an all position 70 ksi SMAW stick electrode. Nearly all of the prequalified weld details in D1.1 are qualified for SMAW. The procedures are qualified for the process. The appropriate filler metal/electrode must be selected for the base metal. Does this answer your question?

A reinforcing fillet is another alternative. But the required size is much larger than 3/16" in order increase the capacity, with the profile of the existing weld. Therefore many more passes are required.

http://www.FerrellEngineering.com

 

[gtaw]

Any weld made one side without backing or back gouging is considered as a partial joint penetration weld per AWS D1.1 unless the procedure is qualified by testing in accordance with Clause 4 (AWS D1.1-2010).

The fact that the joint is considered to be PJP affects the design considerations. The allowable unit stress is essentially reduced for a PJP to that of a fillet weld of the same size. Both the PJP groove weld and the fillet inherently have a crack initiation point where ever incomplete joint penetration exist.

The welding technique employed for CJP joints in tubular structures, with or without backing, have additional requirements imposed by AWS D1.1. For instance, weld starts and stops must not be located at the radiused corners of the tubular sections and must be made continuous around the corners (see note 2, Figure 3.6, D1.1-2010). The root face (land) must be zero per clause 3.13.5, 5th paragraph. And there are additional volumetric inspection requirements for CJP groove welds subject to fatigue ranges B and C if the loading is in tension transverse to the weld axis.

Best regards - Al

 

[Duwe6]

You could get a GOOD UT tech and have him/her measure the as-built, actual penetration. The measured missing groove penetration in your welds can be easily compensated for using a reinforcing fillet weld, as connectegr mentioned.

As fillet weld strength is calc'd through the throat, use 1.4 x 'measured missing penetration' to give the minimum leg size of the reinforcing fillet.

Or save some money -- don't hire the UT tech. Make a mock-up joint using a 1/8" land [it's my preferred size, too] and weld it exactly like you would the production items. Add a generous reinforcing fillet - 1/4" to 5/16". Then saw out several sections thru your weld, polish and etch them [muratic/sulfuric acid works well].

Now show the cross-sectioned weld to your non-welding engineer, and tell him that this proves that you can reliably make welds that develop the full material strength of his [overstrenght] tubesteel column.

 

[RobsVette]

Hey guys, thanks for all the replies so far.

Based on the recommendations here, I went out and did a field inspection of the suspect weld and it turns out the welder put a 7/16" fillet weld over the partial penetration weld.

We submitted this to the engineer who is going to review it and decide if any corrective action needs to be take.

Thanks again for all the replies.

 

[connectegr]

Just curious...

From a cost/time standpoint, is the PJP with reinforcing fillet more economical than providing backing and using the CJP? I am fairly proficient from a technical and design side. But, I am always interested in practical implications of these discussions. Thanks.

http://www.FerrellEngineering.com

 

[Duwe6]

Excellent question, sir. Theoretically, a PJP + fillet is more weldmetal and thus takes more welding time. In practice, they tend to be a 'wash' -- no real difference in wewlding time.

Where the savings comes in is the fitting of the joint. A Full-penn has to be supported by something else during alignment -- root gap has to be maintained. Usually an external attachment "fitup dog" is welded on. Post-weld, the dog has to be removea and the 'dog marks' ground smooth and blended such that archeticts and engineers cannot detect that something 'extra' was added to their structure.

With a PJP that has a 1/8" land, the fitter can just jam the member home, and have the welder tack it in place as soon as the angularity [or plumbness] is correct. Fast fitup saves a lot of time & money.

 

[gtaw]

The only fly in the ointment is the lower allowable stress permitted for fillets and PJP groove welds compared to the high allowable stresses permitted when CJP groove welds are used.

It can be difficult to develop the full strength of some joints (butt joints) when the lower allowable stress permitted for PJP and fillet welds are considered.

Best regards - Al